First Results Obtained Using TEM and CLSM. Part I: Tantulus Larva

Organisms Diversity & Evolution (2018) 18:459–477 https://doi.org/10.1007/s13127-018-0376-4

ORIGINAL ARTICLE

Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva

Alexandra S. Petrunina1 & Jens T. Høeg2 & Gregory A. Kolbasov3

Received: 21 December 2017 /Accepted: 16 August 2018 /Published online: 28 August 2018 # Gesellschaft für Biologische Systematik 2018

Abstract The morphology of Tantulocarida, a group of minutely sized ectoparasitic Crustacea, is described here using for the first time transmission electron microscopy (TEM) and confocal laser scanning microscopy (CLSM). This enabled a detailed analysis of their internal anatomy to a level of detail not possible with previous light microscopic investigations. We studied the infective stage, the tantulus larva, attached on the crustacean host in two species, Arcticotantulus pertzovi and Microdajus tchesunovi, and put special emphasis on cuticular structures, muscles, adhesive glands used in attachment, sensory and nervous systems, and on the organs used in obtaining nutrients from the host. This allowed description of structures that have remained enigmatic or unknown until now. A doubly folded cuticular attachment disc is located at the anterior-ventral part of the cephalon and used for gluingthelarvatothehost surface with a cement substance released under the disc. Four cuticular canals run from a cement gland, located ventrally in the cephalon, and enter an unpaired, cuticular proboscis. The proboscis can be protruded outside through a separate opening above the mouth and is used for releasing cement. An unpaired and anteriorly completely solid cuticular stylet is located centrally in the cephalon and is used only for making a 1-μm-diameter hole in the host cuticle, through which passes a rootlet system used for obtaining nutrients. The rootlet system is a direct extension of the anterior gut of the tantulus, but inside the host, it consists of cuticle only. Several muscular systems seem to degenerate very quickly after the tantulus has settled on the host. The central nervous system is reduced to an absolute minimum in terms of both cell size and number. We discuss the morphology of the tantulus in light of the Tantulocarida being a sister group to the Thecostraca or even nested within that taxon.

Keywords Tantulocarida . Thecostraca . Comparative anatomy . Phylogeny . TEM . CLSM

Introduction mostly in the deep sea (Copepoda, Isopoda, Amphipoda, Ostracoda, Cumacea, Tanaidacea). Described as a class only The Tantulocarida is a group of minutely sized and ectopara- in 1983, tantulocarids have a truly enigmatic morphology and sitic Crustacea, infesting other meiobenthic crustaceans and remain one of the most poorly studied groups of arthropods. A tentative life cycle of the Tantulocarida has been reconstructed based on data gleaned from several different species (Huys * Alexandra S. Petrunina et al. 1993), but they have never been successfully reared in [email protected] the laboratory. The infective larval stage called the tantulus is one of the smallest multicellular organisms, but nevertheless it Jens T. Høeg [email protected] retains a general crustacean body plan (Olesen et al. 2014). This minute stage consists of a cephalon, six thoracic somites Gregory A. Kolbasov with paired natatory thoracopods, one limbless somite, and an [email protected] unsegmented abdomen (Figs. 1cand2). The cephalon lacks 1 Biological Faculty, Department of Invertebrate Zoology, Moscow any appendages, but has an oval-shaped, sucker-like attach- State University, Moscow, Russia ment organ located on its ventral side (Fig. 3) and also pos- 2 Department of Biology, Marine Biology Section, University of sesses a cuticular stylet for penetrating the cuticle of the host. Copenhagen, Copenhagen, Denmark In the currently accepted life cycle, the free-swimming 3 Biological Faculty, White Sea Biological Station, Moscow State tantulus eventually settles on the prospective host (Fig. 1), University, Moscow, Russia pierces the host cuticle, and starts feeding from it by means 460 Petrunina A.S. et al.

Fig. 1 General morphology and rootlet system of the Tantulocarida (CLSM). Parthenogenetic female (a, b)and tantulus larva (c)of Arcticotantulus pertzovi on copepod host Bradya typica. a Parthenogenetic female attached to the host cuticle, general view, dorsally. b Cephalon and anterior part of the egg sac, a rootlet system originating from under the oral disc of the parasite. c Tantulus larva attached to the host cuticle with rootlet system at early stage of development. r.s., rootlet system; st., stylet; u.c., umbilical cord. Scale bars in micrometers

of the rootlet system that enters through the hole (Fig. 1b, c). sexual stages. These form from the attached tantulus, but are The attached tantulus now passes through a complex meta- subsequently are released as free-living, non-feeding males morphosis leading to either a parthenogenetic female or to and females equipped with natatory thoracopods. After a pre- sexual male or female forms. The parthenogenetic female is sumably internal fertilization, the free female releases a new a sac-like stage that remains attached to the tantulus and thus generation of tantulus larvae. While already by itself complex, the host. It contains numerous eggs from which develop a new this hypothetical life cycle (Huys et al. 1993) is also chal- generation of infective tantulus larvae (Fig. 1a). The second lenged by the observation of a very aberrant nauplius stage type of metamorphosis results in a single male or female that does not fit into the currently accepted scheme (P.

Fig. 2 General morphology of the tantulus larva. a Serratotantulus chetoprudae (SEM). b Arcticotantulus pertzovi (TEM, longitudinal section) Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 461

Fig. 3 Cephalon of the tantulus larva: general anatomy on longitudinal cross section. a Schematic drawing, nutrition holes in cuticle of rootlet system indicated by arrows. b TEM (Arcticotantulus pertzovi). a.d.pr., anterior disc of proboscis; br., brain; cem., cement; c.g., cement gland; c.g.c., cement gland cavity; c.g.d., cement gland duct; cut., cuticle; c.v., cuticular villi of the oral disc; ed., epidermis; g.c., gut cell; h.cut., host cuticle; np., neuropil; o.d., oral disc; ov., ovary/germinative cells; pr., proboscis; pr. c. op., proboscis cavity opening; r.s., rootlet system; st., stylet; st.r., stylet retractor. Scale bars in micrometers

Martinez Arbizu, personal communication). It is therefore ap- two other groups of Thecostraca are also either obligatory parent that we still lack a satisfactory understanding of the life parasites (the Ascothoracida) or assumed to be so (the cycle for tantulocarids. Facetotecta), and the facetotectans even have a metamorpho- The tantulus fulfills the same role as the cypridoid larva sis, which rivals the rhizocephalans in complexity (Glenner in the hypothetically closely related Thecostraca by being et al. 2008). Obviously, therefore, it is of high interest to study responsible for attachment and metamorphosis into the en- the tantulus and compare it in detail with the infective stages suing stages. Furthermore, like the kentrogon stage in par- and general adaptation to parasitism in other Thecostraca. But asitic rhizocephalan cirripedes, the tantulus possesses a cu- unlike rhizocephalans, where the settlement and host infection ticular stylet, passes through a complex metamorphosis, has been studied in minute detail with transmission electron and the resulting parasite derives nutrients by means of a microscopy, the tantulus and all other stages of the system of rootlets inside the crustacean host. On the other tantulocarid life cycle has until now been examined by light hand, the rhizocephalan parasite is first injected as an en- microscopy only. These studies have revealed a lot of struc- tirely endoparasitic stage that only later emerges on the tural details, but they also raise many questions and are in surface of the host, while the tantulocarid parasite remains general very difficult to compare with TEM level information. external except for the rootlets. A very preliminary TEM study of the tantulus was given by The apparently close relationship between the Petrunina et al. (2014). The purpose of this paper is therefore Tantulocarida and the Thecostraca and the similarities be- to provide a comprehensive description of the anatomy of the tween the tantulus and the rhizocephalan kentrogon obviously tantulus larva using both TEM and CLSM techniques. Based raise the question whether the two groups could be closely on this, we will compare the results with previous light micro- related or whether tantulocarid parasitism evolved scopic accounts and thereby provide a platform for discussing convergently. It adds to the complexity of this problem that tantulocarid relationships and evolution. 462 Petrunina A.S. et al.

Materials and methods dorsally in one specimen (Fig. 5e) and by position could possibly have belonged to a stylet retractor (Fig. 3a). The cuticle varies in Two species of the Tantulocarida Arcticotantulus pertzovi thickness from 0.20 to 0.70 μm and is clearly differentiated into a Kornev, Tchesunov, Rybnikov, 2004 and Microdajus thin electron-dense and 20-nm-thick epicuticle and homogenous tchesunovi Kolbasov et Savchenko, 2010 are known to para- procuticle (Figs. 5f, g, 3,and4). Procuticle can consist of up to sitize a harpacticoid Bradya typica Boeck, 1873 and a tanaid five irregular layers of different density (Fig. 5g). Typhlotanais sp. that are common meiobenthic crustaceans inhabiting silty sediments of the Kandalaksha Bay in the vi- Thoracic somites and muscular system cinity of the White Sea Biological Station (66°31′41″ N, 33°11′08″ E). Bottom mud was obtained from the depths of Most of the thorax and the thoracopods are filled with a pop- 20–40 m using light hyperbenthic dredge. A standard bub- ulation of small-sized cells with a high nucleus-to-cytoplasm bling or flotation method was applied to extract meiobenthic ratio. The nuclei contain highly condensed chromatin and crustaceans from the silt (Higgins 1964). Samples were then have almost polygonal shapes with a largest diameter of about inspected under a stereo microscope to get tanaids and ca. 1.5 μm and with very little surrounding cytoplasm. We harpacticoids infested with the Tantulocarida. Specimens at cannot presently allocate any function to these cells. different stages of development attached to their hosts were Surprisingly, no clear muscles are present in the attached fixed for electron and confocal microscopy. tantulus larvae of A. pertzovi (Fig. 6a, c) and M. tchesunovi For transmission electron microscopy several specimens of investigated here, except traces of retractor muscles of stylet settled and metamorphosing tantulus larvae of A. pertzovi and (Fig. 5e) and degenerated transverse flexor muscles in first M. tchesunovi were fixed in 2.5% glutaraldehyde on 0.1 M thoracic somite (Fig. 6b). In addition, we observed in some PBS. The samples were then washed three times for 15 min in specimens the attachment sites of degenerated muscle strands PBS and stored at 4 °C before being post-fixed in buffered 1% (stylet retractor, thorax and limb flexors), identified by their

OsO4, dehydrated through ethanol series, and embedded in very characteristic electron-dense semi-desmosome contacts Araldite resin. Ultrathin sections were cut with a diamond (Figs. 3aand6b, d). knife using Ultracut-R Leica ultratome and stained for 40 min with saturated uranyl acetate and for 5 min with lead Oral disc citrate. The sections were examined and photographed with a JEOL JEM 100B and JEOl-1011 80 kV. The attachment apparatus in the form of an oval-shaped (ca. For scanning electron microscopy, a settled tantulus larva 25 μmlongbyand15μm wide) oral disc is situated anterio- of Serratotantulus chertoprudae Savchenko and Kolbasov, ventrally on the cephalon (Figs. 3 and 4). The oral disc is

2009 was postfixed in 1% OsO4, dehydrated in a graded eth- partially covered by anterior and lateral cuticular projections, anol series and acetone, and critical point dried in CO2.The 0.17–0.70 μm thick, while the disc cuticle itself is but 0.13– specimen was then sputter-coated with platinum–palladium 0.19 μm thick (Fig. 7a, c). The ventral surface is covered with alloy and examined in a JEOL JSM-6380LA microscope at numerous cuticular folds and villi (Fig. 7d–f). It and the whole operating voltages of 15–20 kV. disc structure is connected to the ventral side of the cephalon Several specimens of A. pertzovi attached to their hosts were by a short cuticular neck, ca. 3.2 μm in diameter and with fixedwith4%paraformaldehydein1MPBSfor4h,washedin thick walls enclosing the anterior part of the gut (Fig. 7b, c). 1 M PBS, and stained with Congo Red to be observed with Paired cavities (up to 2 μm high) are seen inside the cuticle of TCS CP5 Leica confocal laser scanning microscope. the oral disc (Fig. 7c, f), but it is not clear whether they are Photo plates and line drawings were prepared in Inkscape 0.92. connected with general body cavity or just represent gaps between cuticle layers. The mouth opening is located centrally on the oral disc (Figs. 3 and 7a, arrowhead) and anteriorly to Results this a separate opening for the proboscis (Figs. 3aand4a). In A. pertzovi, the entire ventral side of cephalon has numerous Epidermis and cuticle longitudinal cuticular ridges composed of both epi- and procuticle (Figs. 5hand7b). The epidermis consists of a 2–3-μm-thick layer of epithelial cells (Figs. 3 and 4) with comparatively large elongated nuclei con- Cement taining prominent nucleoli (Fig. 5a–f). They also contain endoplasmic reticulum indicative of a high synthetic activity Until the system of rootlets grows out into the host, the attach- (Fig. 5d). In several specimens, there are multi-membranous ment is ensured entirely by a cement substance presumably of bodies of unknown function in the ventral epidermis (Fig. 5b, protein nature that is secreted under the oral disc and serves to c). An attachment site for paired muscle strands was observed glue the parasite to the cuticle of the host (Figs. 3, 4,and7c–f). Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 463

Fig. 4 Cephalon of the tantulus larva: general anatomy on frontal cross- cement gland duct; ed., epidermis; g.c., gut cell; o.d., oral disc; ov., ovary/ section. a Schematic drawing. b TEM (Microdajus tchesunovi). a.d.pr., germinative cells; pr., proboscis; pr. cav., proboscis cavity; pr. c. op., anterior disc of proboscis; br., brain; cem., cement; cut., cuticle; c.g.d., proboscis cavity opening; st., stylet. Scale bars in micrometers

The investigated species differed with respect to the type of widens into a discoid shape (Figs. 3, 4,and8). The structure of cement as seen in the TEM pictures. In A. pertzovi, the cement cement canals is quite distinct, having very thin walls of epicu- is fibrous, consisting of more or less loosely distributed fibers, ticle and are in A. pertzovi reinforced by ring-shaped thicken- 15–45 nm thick (Fig. 7d, e). In M. tchesunovi, the cement ings, rendering them corrugated (Figs. 3 and 8a, b, e, f). The consists of a homogeneous substance of medium electron den- ducts are 0.3–0.5 μm in diameter in the anterior part (Fig. 8e, f), sity (Fig. 7c, f). This difference between the two species was but decreasing slightly toward the posterior end of the cephalon. seen both in the cement of newly attached tantuli and in that of In M. tchesunovi, the duct walls have a smooth 0.8-μm-thick already metamorphosed specimens. In A. pertzovi and cuticle without any ring structures (Fig. 9a, b). M. tchesunovi, the coverage of cement is confined to the space underneath the attachment apparatus (Fig. 7c, d), while in Serratotantulus chertoprudae, small portions of the glue ap- Cement glands pear to be squeezed out (Savchenko and Kolbasov 2009). In detached specimens of M. tchesunovi, the cement can be seen From the proboscis cement gland, ducts run as a group of four to fill the entire volume under the oral disc, having smooth backwards into the dorsal part of the cephalon (Fig. 3). In front surface that replicates the cuticle of a host (Fig. 7a). of the brain, they split in two pairs and then proceed ventrally (Figs. 4 and 9a), while in A. pertzovi also losing the corrugated wall structure. For the major part of their course, there is no Proboscis epithelium beneath the cuticle of the ducts. Ultimately, the ducts terminate in four large and cuticle-lined cavities located ventral- The proboscis is a funnel-shaped, ca. 6-μm-long cuticular ly in the cephalon (Figs. 3 and 9c–e). These cavities have a structure, situated anteriorly in the cephalon and lying longi- thicker cuticle than in the ducts proper (Fig. 9d), and they are tudinally above the digestive tract (Figs. 3 and 8a–c, e). either spherically shaped (Fig. 9e) or resemble a deflated bal- Having thin epicuticular walls (Fig. 8e, f), it lies in its own loon (Fig. 9c, d). Both the terminal part of the ducts and the cavity and resembles the finger of a glove intruded into the terminal cavities lie within what appears to be epithelium cells, cephalon (Fig. 8a). The opening of this cavity, observed here and it even seems as if each duct and its particular cavity is for the first time, lies anteriorly to the mouth (Figs. 3aand4a). associated with one single such cell (Fig. 9b–e) that again forms Four cuticular ducts from the cement gland enter the probos- part of the cement gland (Figs. 3 and 9c, e, f). The ventrally cis and terminate with four small pores on its distal part which located cement gland consists of secretory cells, having big 464 Petrunina A.S. et al.

Fig. 5 Epidermis and cuticle, tantulus larva (TEM). a, g Microdajus tchesunovi; b–f, h Arcticotantulus pertzovi. a Epidermis. b Transverse cross- section through cephalon, ventrally. Multimembranous body in epidermis indicated with astar.c Longitudinal section through cephalon, multimembranous bodies in epidermis indicated with a star, ventral part. d Transverse cross- section through cephalon, epidermal cell with endoplasmic reticulum, ventral part. e Transverse cross-section through cephalon showing attachment site of the paired muscle strands to the dorsal body wall. f Metamorphosing tantulus larva, transverse section through the middle part of cephalon showing cuticle and epithelium structure. g Tantulus larva, sagittal section through the middle of cephalon showing cuticle and epidermis structure. h Metamorphosing tantulus larva, ventral margin of cephalon with cuticular ridges, anterior part, transverse section, TEM. ed., epidermis; e.r., endoplasmic reticulum; g., gut; m.a., muscle attachment site; n., nucleus; nu., nucleolus; procut., procuticle; v.r., ventral cuticular ridges. Scale bars in micrometers

nuclei with distinct nucleoli and a cytoplasm with copious en- in the distal part to 0.3–0.5 μm in the proximal part (Fig. 10b– doplasmic reticulum and as numerous vesicles (Fig. 9c–f). d). In M. tchesunovi, the basal part of the stylet expands dorsally and curves forward ventrally (Fig. 10a). Stylet Rootlet system and gut The stylet is an unpaired, cuticular organ located longitudinal- ly in the middle of cephalon (Figs. 1b, 3a, 4,and10a), where it The free-swimming tantulus larva is a non-feeding stage, but is already fully developed in the free-swimming tantulus soon after settlement it turns into an ectoparasite that feeds by (Huys 1991). In transverse section, it has a roundish shape means of a system of rootlets inside the host tissues. Newly (Fig. 7b, d) changing to triangular near the proximal end. attached tantuli have no rootlets. We could not observe the The diameter is 0.50–0.8 μm at the tip but increasing to ca. rootlet formation directly, but in attached tantuli assumed to 1.8 μmatthebasalmuscleattachmentsite(Fig.10c). It is be slightly older, the anterior and cuticular part of the gut has hollow for most of its length with the cavity opening proxi- sent an outgrowth into the host through the hole produced by mally completely (Fig. 10a–c), but is completely solid at the the stylet (Figs. 1b, c and 3). The form of the resulting rootlet distal tip (Figs. 3 and 10d, e). The stylet cuticle consists of an system is variable, but seems normally to consist of main trunk outermost 20 nm epicuticle and a procuticle without any dis- that can issue branches, perhaps depending on the age of the tinct fibrous structure and ranging in width from 0.15–0.2 μm parasite (Figs. 1band11). The walls of the rootlet consists of a Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 465

Fig. 6 Internal structure of the thoracic segments, tantulus larva, TEM (Arcticotantulus pertzovi). a, c Longitudinal cross-sections through whole body and first and second thoracic segments. b, d Transverse cross-sections through first thoracic segment with traces of dorsoventral muscles and muscle attachment sites, ventral part enlarged in (d). ceph., cephalon; d. m., disintegrating muscles; m.a., muscle attachment site; thp., thoracopods. Scale bars in micrometers

very thin cuticle (60–140 nm), in which both a procuticle and consists of cuticle only (Fig. 12a–c, f). Further posteriorly, the an epicuticle can nevertheless be distinguished (Fig. 11b, c, e). gut walls are formed by cells with numerous very thin out- Surprisingly, there seems to be no underlying epidermis or any growths (Figs. 3, 4,and12e, f) that occupy the entire central other cellular elements inside the rootlets, which therefore ap- space The gut is entirely cephalic since its posterior part adjoins pears to be entirely cuticular. Newly attached tantuli have not the brain (Fig. 12c, f), and this corresponds with absence of an yet developed the rootlet system, but could nevertheless have anus in the tantulus larva. The anterior gut lumen is filled with material inside their gut cavity, indicating they can absorb nu- hemolymph of the host even before a rootlet system has devel- trients through the mouth opening which is oppressed to the oped inside the host (Figs. 11dand12e, f). puncture made by stylet (Figs. 11dand12f, g). At the mouth opening in the central part of the oral disc, the Brain rootlet system is directly continuous with the anteriormost and entirely cuticular part of the gut (Fig. 11a). From the mouth, the The bi-lobed and dumbbell-shaped brain is located in the pos- gut then continues inside a cuticular thickening of the oral disc terior part of cephalon (Figs. 3, 4,and13a, c, d), measuring (neck region) (Figs. 3 and 12c, f). Further inside the cephalon, maximally 12 μminlengthlong,20μm in width, and 13 μm the gut first retains a thick cuticular wall (Fig. 12a), but this is in height and with no evident divisions into a proto-, deuto-, eventually replaced by very thin epicuticle (Figs. 3, 4,and12b, and tritocerebrum. The particular form and location of the c, f). In this region, the gut cuticle lies above an epithelium brain depends on both the species and the future development consisting of very large (4–6 μm) cells with big nuclei (about pathway taken by the larva. In a tantulus destined to metamor- 2 μm) and containing fibrinous chromatin evenly spread inside phose into a parthenogenetic female, the ovary develops in the a round-shaped nucleus (Figs. 3, 4,and12c, d, f). The lumen of posteriorpartofthecephalon(Figs.3, 4,and13c), thus the gut is only distinct in the anterior part, where the wall shifting the brain slightly toward the anterior part. The total 466 Petrunina A.S. et al.

Fig. 7 Attachment apparatus of the tantulus larva (a, c, e Microdajus part of cephalon, TEM. d Tantulus larva, anteriormost part of cephalon, tchesunovi; b, d, e Arcticotantulus pertzovi). a Metamorphosing longitudinal section, TEM. e Transverse section of the oral disc showing tantulus larva. Ventral side of the cephalon with oral disc covered with cement structure, TEM. f Transverse section of the anteriormost part of cement, mouth opening on the disc marked with an arrowhead, SEM. b cephalon with oral disc showing cuticular villi, TEM. cem., cement; c. v., Metamorphosing tantulus larva, anterior part of cephalon, oral disc cuticular villi of the oral disc; ep., epicuticle; g., gut; h.c., host cuticle; removed showing cuticular neck of the oral disc with the gut centrally, n.d., neck of the oral disc; o.d., oral disc; o.d.cav., cavity in oral disc; p.c., SEM. c Newly attached tantulus larva, transverse section of the anterior projections of cephalon. Scale bars in micrometers number of pericarya is estimated to be approximately 220– tantulus larva has 11 pairs of pores (AI–AIV,DI–DIV,LI–LIII). 250. They are very small (slightly more than 1 μm) and lie In M. tchesunovi, the dorsal pores are shifted to the hind margin close together as a cortical layer surrounding distinct, ca. and the openings of BD^ pores face in posterior direction. 4-μm-wide neuropil. The cortex of cell bodies is 6–7 cells Simple pores consist of two parts. An outer pore chamber thick in the mid-dorsal part and 2–3 cells thick laterally and about 0.6 μm in diameter and 0.5–1 μmdeep(Fig.14e, k–m) (Fig. 13d, e) with the large ca. 1-μm-wide nuclei occupying located directly in the cuticle and an inner cavity 0.3–0.5 μm the major part of the cell volume, leaving only a thin layer of wide and 0.5–1 μm deep located in the epidermis (Fig. 14f, k, cytoplasm (Fig. 13d, f, g). Masses of highly compact, elec- l). The inner chamber is in contact with a sensory dendritic tron-dense, and homogeneous chromatin are mainly concen- process. The walls of the pore, both in the outer and inner trated along the perimeter of the nuclei (Fig. 13f). Electron- chambers, are outlined with thin epicuticle, while the con- dense vesicles indicative of synapses can be seen in the neurite stricted and ca. 200-nm-wide communication between the fibers of the neuropil (Fig. 13a, b). There are no clear neuro- two compartments is reinforced by a neck of thicker cuticle glial cells adjoining the cell bodies and no distinct enveloping (Fig. 14f, k, l). The sensillate pores/pits lack both cuticular sheath (Fig. 13a, c, d, g). chambers and canal and the dendritic process is directly connected with the minutely sized sensilla (Fig. 14d, e). Sensory structures Ovary The tantulus larva has neither eyes nor sensilla-carrying appendages and the only sensory organs are a number of paired cephal- In tantulus larvae presumed to be metamorphosing into parthe- ic pores/pits (Fig. 14) with or without sensilla (Fig. 14a, b, g). nogenetic females, the primordial ovary is located in the rear Using the pore formula of Boxshall and Vader (1993), part of cephalon (Figs. 3, 4, 13c, and 15a). It is a group of large M. tchesunovi has four pairs of pores distributed on the cephalon cells forming a compact rosette-like structure and all having the according to the formula AI,DI,DII,LI, while in A. pertzovi same appearance: (1) high nucleo-cytoplasmic ratio; (2) large Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 467

Fig. 8 Proboscis and cement gland ducts, TEM. Arcticotantulus pertzovi (a–c, e, f), Microdajus tchesunovi (d). a Tantulus, longitudinal section through the midline of cephalon, anterior part. b Tantulus, cement gland ducts inside proboscis, transverse cross-section. c Tantulus, anterior disc of proboscis, longitudinal section, openings of the cement gland ducts marked with arrowheads. d Cement released through the anterior disc of proboscis under the oral disc, transverse cross- section, opening of the cement gland duct marked with an arrowhead. e Metamorphosing parthenogenetic female, cement gland ducts entering proboscis, transverse section through anterior part of cephalon. f Tantulus, cement gland ducts with and without cement, longitudinal cross-section. a.d.pr., anterior disc of proboscis; cem., cement; c.g.d., cement gland duct; cut., cuticle; g., gut; h. cut., host cuticle; o.d., oral disc; pr., proboscis; pr. cav., proboscis cavity. Scale bars in micrometers

nuclei (about 2.2 μm) with diffuse chromatin; (3) large nucleoli of this extremely small stage and correlate with findings with (0.7 μm) (Fig. 10a, b). The cytoplasm contains small, electron- previous studies based on light microscopy or SEM only. With dense areas that may possibly be referred to as nuage (Fig. 15a, respect to several organ systems, notably the rootlet system b, arrowheads), but we could not find that these dark spots are and the gut and the cement apparatus, we are for the first time associated with mitochondria. Since all these characters are able to provide a comprehensive account that allows both typical of germ cells, we consider this organ as the ovary of discussion and comparison with other taxa putatively related the parthenogenetic stage, which is then removed from the to the Tantulocarida and their larval stages, most notably the cephalon to the growing cuticular sac (Fig. 1a). cypris and kentrogon stages of the Cirripedia Rhizocephala.

Attachment apparatus Discussion The attachment process The cephalon of tantulus is devoid of We have for the first time provided a full account of the paired appendages, so the proboscis and the associated cement tantulus larva of the Tantulocarida by means of transmission glands are wholly responsible for attachment of the tantulus electron microscopy and supplemented by information using larva which is ensured entirely by chemical means. Previous CLSM. This allows us to describe in detail the organ systems SEM studies have shown that the free-swimming tantulus can 468 Petrunina A.S. et al.

Fig. 9 Cement glands, ducts, and cuticular cavities of cement glands (TEM). Microdajus tchesunovi (a, b), tantulus larva, frontal cross-section through central part of cephalon; Arcticotantulus pertzovi (c–f), tantulus larva, transverse cross- section of cephalon through cement glands. a Two pairs of cement gland ducts running along the gut. b Cement gland ducts. c Right part of cement gland (marked with dashed outline) with cuticular cavity and cement gland duct fusing with it. d Cuticular cavity of the cement gland. e Transverse cross-section of cephalon through cement glands (marked with dashed outline) and cuticular cavities. f Transverse cross-section through cement gland. c.g., cement gland; c.g.c., cement gland cavity; c.g.d., cement gland duct; e.r., endoplasmic reticulum; g., gut; n., nucleus; nu., nucleolus. Scale bars in micrometers

protrude the proboscis out of the cephalon (Huys et al. 1992), that probably were associated with either the proboscis or the but the true position of proboscis cavity and its opening ante- stylet retractor muscles. In Dicrotrichura tricincta Huys 1989, riorly to the mouth was observed here for the first time proboscis was connected with muscular strands running fur- (Figs. 3aand4a). After initial cement secretion, the proboscis ther inside cephalon (Huys 1989), and it was suggested that must be withdrawn back into its cavity thus pulling the larva the so-called striated organ described in other species toward the host. This is followed by release of the rest of the (Boxshall and Lincoln 1983) is simply a proboscis retractor. cement that can now spread under the entire attachment disc and provide a firm attachment to the surface. The cement In the tantulus, the cement is secreted only under It is somewhat baffling that we could not find any muscles the oral disc and the adhesion depends upon the chemical prop- responsible for either the protraction or retraction of the pro- erties of this substance. The special morphology of the attach- boscis. This must be explained by a very rapid degradation of ment organ reveals several adaptations to an efficient adhesion. myofibers as soon as they have completed their function as The cuticular outgrowths of the cephalon around the oral disc suggested by Boxshall (1991). In agreement with this, we did may well serve to prevent spilling out of cement and the folds in one specimen find clear muscle attachment sites (Fig. 5e) and villi on the disc itself will increase the surface area, thus Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 469

Fig. 10 Stylet of the Tantulocarida (TEM, a–d;SEM,e). Microdajus transverse section of stylet, middle part. c Tantulus, transverse section of tchesunovi (a, d), Arcticotantulus pertzovi (b, c), Serratotantulus stylet, basal part. d Tantulus, sagittal section of stylet, distal part. e Solid chertoprudae (e). a Metamorphosing tantulus, longitudinal section tip of stylet surrounded by anterior part of gut. ep., epicuticle; g., gut; m.s., through the midline of cephalon, showing position of stylet. b Tantulus, muscle strand; st., stylet. Scale bars in micrometers also serving for firm fixation. Functional similarities to this may studied by us had already started to degrade. Exactly the same again be found in the structure of the antennular attachment is seen in kentrogons of Lernaeodiscus porcellanae Muller, organ in cirripede cyprids (Bielecki et al. 2009). 1862 when their cement secretion ceases a few hours after first attachment. Correlation between light microscopy and TEM The cement gland The adhesive glue itself is produced in data is always fraught with difficulty, but we suggest that so- paired cement glands and is transported through four cuticular called globular structures and glandular bodies described by ducts to the distal end of the proboscis. We suggest that prior several authors from both live and fixed tantuli (Huys 1991; to secretion, the cement is stored in the cavities associated Huys et al. 1992) correspond to the cuticular cavities were with the proximal end of each duct. The actual secretory cells here found to be associated with the cement glands. of the gland have all the characteristics of high synthetic activity, and we surmise that the secreted substance is at least in Digestive system part of a proteinaceous nature. The amount needed for adhesion is very small and released immediately at attachment, so Stylet The tantulus larva starts feeding only after settlement on we assume that the cement gland in the settled specimens host, when it becomes a true parasite. The mechanism of 470 Petrunina A.S. et al.

Fig. 11 Rootlet system of the Tantulocarida (TEM). Microdajus tchesunovi (a, e), Arcticotantulus pertzovi (b–d). a Metamorphosing tantulus, transverse section through anterior part of cephalon showing connection of rootlet system with anterior gut. b Metamorphosing tantulus, rootlet system inside host tissues, transverse section. c Enlarged cuticular wall of the rootlet system. d Newly attached tantulus (rootlet system is not developed yet), longitudinal section through anteriormost part of cephalon medially, stylet puncture in host cuticle indicated with arrowhead. e Metamorphosing tantulus, rootlet system (filled with host content— hemolymph) in host tissues, transverse section. cem., cement; ep., epicuticle; g., gut; h.c., host cuticle; o.d., oral disc; p.c., projections of cephalon; r.s., rootlet system. Scale bars in micrometers

obtaining nutrition was developed Bde novo^ as an adaptation to muscular attachment sites that most likely relate to both stylet the parasitic style of life, and thus it has several unique features. and proboscis muscles. In Tantulacus hoegi Huys, Andersen, An unpaired cuticular stylet is used to get access to host fluids Kristensen, 1992, the stylet is claimed to have a thin duct with through a single puncture made by it. The particular mechanism a distal opening armed with a row of barbs and also to be asso- of the stylet movement is not yet fully understood, but protractor ciated with paired glandular bodies at the basis of stylet, sug- and retractor muscles were described in detail in free-swimming gested to secrete a substance for dissolving the host cuticle larva using light microscopy (Huys 1991; Huys et al. 1992). The (Huys et al. 1992). In contrast, we found that the stylet always protractors lie laterally on both sides of stylet and go from the end had a solid tip and without any associated glands, so at least for of the so-called gubernaculum to the posterior barbs of stylet. the three species investigated here, it is clear that the stylet is used Paired retractors originate from the basal part of stylet and are only for mechanical puncture of host integument. attached dorsally to the cuticle of posterior part of cephalon. We could not observe any of these muscles in our attached speci- Rootlet system The opening left by the stylet is so small that mens, so presumably they start to disintegrate immediately after the tantulus could not efficiently obtain nutrition from the host they have completed their function. Furthermore, we did find without some sort of a pumping mechanism for which we Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 471

Fig. 12 Gut anatomy, tantulus larva (TEM). Arcticotantulus pertzovi (a–b, e–g), Microdajus tchesunovi (c, d). a Anterior part of cephalon, general view, transverse cross-section showing anterior gut with thick cuticular dorsal wall and distal part of proboscis. b Anterior part of cephalon, transverse cross- section, gut with thin cuticular lining (indicated by arrowheads). c Frontal cross-section through cephalon, general view of the gut (gut wall indicated by arrowhead) with enlarged part showing the gut cell (d). e–g Longitudinal cross-section through middle line of cephalon (f) with enlarged anteriormost (g) and posterior (e) parts of gut showing host hemolymph inside the gut. br. c., brain cell; c.g.d., cement gland duct; g., gut; g.c., gut cell; h.c., host cuticle; h.h., host hemolymph; pr., proboscis; st., stylet. Scale bars in micrometers

found no evidence. We can now for the first time prove that It is not yet evident how the formation of the rootlet system the rootlet system is in direct communication with the anterior is performed since in our specimens there were no epidermis cuticular part of the gut. The rootlets must therefore be the cells entering the host tissue that could secrete the rootlet cuticle organ by which the parasite feeds from the host, but the details and enable its expansion into a branched system. Although of this function remain obscure. Unlike the well-described purely speculative, it is possible that such cells are initially rootlet system of rhizocephalan barnacles (Bresciani and present but then withdrawn after cuticle secretion as is known Høeg 2001), the rootlet cuticle of tantulocarids is not to occur in appendages of some arthropods (Høeg 1985). underlined with any epithelial cells that could function in absorption of nutrients. It is probable that transportation of host The gut The cuticular anterior gut forming a guiding tube for juices simply enters the rootlets through tiny holes in the cu- the stylet has been reported for several species (Huys et al. ticle (Fig. 3, arrows), but this still leaves the question how an 1992; Savchenko and Kolbasov 2009), but even based on the efficient flow into the tantulus body is maintained. Possibly, present TEM data, the structure of the middle gut remains the tantulus is so minute that simple diffusion through the uncertain. A tantulus larva already fixed to host does not have rootlets and into the gut is in fact the active mechanism. any evident middle gut with a lumen and microvillus-covered 472 Petrunina A.S. et al.

Fig. 13 Brain anatomy, tantulus larva (TEM). Arcticotantulus pertzovi (a, b, d–g), Microdajus tchesunovi (c). a Longitudinal cross-section through middle line of cephalon, brain indicated by dotted outline. b Enlarged part of brain with electron-dense vesicles containing mediator (indicated with arrowhead). c Frontal cross- section through cephalon, brain indicated by dotted outline. d–g Transverse cross-section through cephalon. d Posterior part of brain with neuropil and bi-lobed cortex layer. e Neuropil, enlarged part of (d). f Neurite cell bodies, enlarged part of (d). g Anterior part of brain located dorsally in the cephalon, transverse cross-section. g., gut; np., neuropil; ov., ovary; st., stylet. Scale bars in micrometers

epithelial walls neither inside cephalon (Fig. 12) nor in the using light microscopy. No additional structures enclosing thoracic segments (Fig. 6a–c). Still, the cells here observed gut and proboscis were detected. However, the features in in the posteriormost part of the gut must in some way be the the complex cuticular neck region of the oral disc may ones responsible for absorption even if they do not resemble a have been what was described as buccal gubernaculum. traditional gut epithelium. Additional cuticular structures such as the so called buc- Cuticle and molting cal gubernaculum and buccal capsule were described by Huys (1991) in the free tantulus larvae of Xenualytus The metamorphosis of the tantulus into a new stage takes scotophilus Huys, 1991 and Campiloxyphos dineti Huys, place without any normal molting process, which begs the 1990 (Huys 1990). In our study, even in newly attached question how the cuticle can expand in area. The cuticle tantuli, there was no clearly discernible such gubernaculum sac with parthenogenetic eggs grows directly from the pos- or buccal capsule. The thick cuticle in the anterior part of terior margin of cephalon after the trunk somites are shed the gut serves as guiding tube for the stylet, and this may away and can expand up to ten times the size of tantulus well have been what was described as a buccal capsule larva. We found that the cuticle of the egg sac of the Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 473

Fig. 14 Cephalic sensory pores, tantulus larva (a–c, g–i—SEM; d–f, j–m— innercavityofDII pore. g–i Posterior margin of cephalon showing simple and TEM). Microdajus tchesunovi (a–f), Arcticotantulus pertzovi (g–m). a sensillate pores. j Transverse section through AIII pore with outer pore chamber Anterior part of cephalon, dorsolateral view. b, c Posterior part of cephalon, and inner cavity. k, l Posterior margin of cephalon, transverse section through dorsal view. d Anterior part of cephalon, longitudinal section through AI pore. DII and DI pores. m Posterior margin of cephalon, transverse section through e Posterior part of cephalon, longitudinal section through DI pore. f Posterior DIV pore. A, D, L, cephalic pores; cut., cuticle; i.c., inner cavity; p.ch., pore part of cephalon, transverse section through pore chamber and beginning of chamber; se., sensilla. Scale bars in micrometers parthenogenetic stage has the same structure as in the unexplained expansion of cuticle area without any molting cephalon of the tantulus larva so a simple stretching cannot occurs during the growth of rhizocephalan externae (Høeg explain the growth of the cuticle. A similar and similarly 1982; Høeg and Lützen 1995). 474 Petrunina A.S. et al.

Fig. 15 Ovary (TEM, tantulus larva). Microdajus tchesunovi, frontal oocytes (nuages in cytoplasm indicated by arrowheads). ed., epidermis; cross-section through cephalon. a Ovary located in posterior part of n., nucleus; nu., nucleolus. Scale bars in micrometers cephalon (nuage in cytoplasm indicated by arrowheads). b Developing

Nervous system and sensory structures Comparison of Tantulocarida and Thecostraca

The nervous system is greatly influenced by the significant The present TEM investigations of the Tantulocarida give miniaturization of the tantulocarid larva. However, all us for the first time a chance for comparison of their inter- modifications have their limits. For example, the size of nal structures with other Crustacea in search of possible neurite bodies is close to its limit of 1 μm. The number homologies. It has for a long time been suggested that of nervous cells in the cortex layer is very low in compar- Tantulocarida could be closely related to the Thecostraca, ison with closely related normal size Thecostraca. and recent molecular phylogenetic studies support this hy- Neurons’ ultrastructure shows typical patterns of miniatur- pothesis and even suggest that they could be nested within ization, such as highly condensed chromatin concentrated this taxon (Khodami et al. 2017; Petrunina et al. 2014). The along the nuclei walls, reduced size of neuron bodies, Thecostraca comprise the assumedly parasitic Facetotecta, tightly packed nervous cells in the cortex layer of the brain, the parasitic Ascothoracida, and the Cirripedia that com- and reduced neuropil, features also observed in other min- prise both parasitic and filter-feeding forms (Grygier 1987; iature arthropods (Makarova and Polilov 2013). Høeg et al. 2009). Although adult thecostracans differ The tantulus must obviously be able to identify the right widely in morphology and life style, their monophyly has host species upon contact and before it commits itself for until now been unchallenged based on very strong evi- irreversible attachment, and it seems that the only structures dence from both larval morphology and molecular data available for this are the putative sensory pores and pits iden- (Pérez-Losada et al. 2004;Høegetal.2009). Whether or tified here by TEM. The inner structure of such cuticular pores not tantulocarids are nested within Thecostraca, there are was studied in several crustacean groups by Halcrow (1993), numerous spectacular similarities between the two taxa and the pores of the tantulus conforms to a general pattern by that need to be discussed in order to decide whether or having a cuticular chamber and an inner epidermal canal or not they represent homologies or convergent evolution: lumen and by being connected with sensory dendritic process- the general life cycle including permanently attached adult es. Somewhat surprisingly, we did not find any signs of a stages, the general tagmosis and function of the larval paraciliary segment that normally is associated with most stages, the mode of fixation on a substratum and the sen- chemo- and mechanoreceptors in crustaceans (Hallberg and sory structures used in this process, and feeding mecha- Skog 2011). It is possible that the morphology of the sensory nisms. Finally, the presence of several parasitic groups pores seen in the tantulus is the result of the extreme minia- within Thecostraca obviously raises the question: are turization of the body. In several species, the free-swimming Tantulocarids closely related to one of these (Høeg et al. tantulus larvae have up to four pairs of cuticular filaments on 2009; Pérez-Losada et al. 2009)? Only based on such a the oral disc (Huys 1991;Huysetal.1992; Martinez Arbizu comparison together with additional molecular evidence and Petrunina 2017), but our TEM study could not prove a can it be decided if the concept of Thecostraca and its sensory function for these structures. intrinsic phylogeny need to be revised. Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 475

General tagmosis and function of the larva despite these differences, a homology cannot be excluded and it would be highly interesting to also compare the chem- The tantulus larva has the same overall tagmosis and mode of ical properties of the cement in these two taxa in search of life as found in the cypridoid larva common to all Thecostraca. possible similarities. There are in fact interesting similarities They all have a short free-swimming life followed by settle- between the attachment process in the tantulus and at least ment on a substratum leading to the metamorphosis into either some of the rhizocephalan cirripedes. In both tantuli and parasites or filter-feeding forms. Like the cypris, the tantulus kentrogons of L. porcellanae, the larva is at first loosely at- also has six pairs of swimming legs, but more evident homol- tached by either proboscis or antennules but is then pulled ogies cannot easily be found. All thecostracans have a toward the substratum and firmly glued by renewed cement cypridoid larvae sharing a suite of unique characters including secretion (Høeg 1985). While this is probably just a (1) six thoracic segments with natatory thoracopods, (2) fron- convergently evolved trait, it nevertheless underlines the need tal filaments associated with compound eyes, (3) compound in both tantulus and cypris/kentrogon to assure a firm attach- eyes with tripartite crystalline cones, (4) lattice organs in a ment for the penetration into the host by means of a stylet. carapace, and (5) prehensile antennules used in attachment by either mechanical or chemical means (Høeg et al. 2004, Lattice organs 2009). None of these features except the thoracopods are present in the tantulus, but this could of course be due to secondary To perform their function in locating potential host or surface modification and loss. Although surprisingly stereotyped, the for the settlement, both cypridoid thecostracan larvae and a cirripede cyprid can also lose structures through specializa- tantulus larva need sensory apparatus. In thecostracan cypridoid tion, with perhaps the most spectacular example being the larvae, the most important sensory organs, aside from eyes, are total loss of thoracopods and even the entire thorax in some the sensilla on the antennules and the specialized chemosensory forms (Kolbasov and Høeg 2007). structures in the carapace called lattice organs and considered as a synapomorphic feature for the Thecostraca (Jensen et al. Life cycle 1994;Høegetal.1998; Høeg and Kolbasov 2002). The tantulocarid larva lacks eyes and both antennules and lattice Tantulocarids are specialized parasites having a complex life organs, and this leaves the sensory pores and pits as the only cycle including both sexual and parthenogenetic phases, where- possible organs responsible for substratum recognition. as other Thecostraca only have a sexual phase. The tantulus The sensory pores studied here have some resemblance to larva could be evolved as an adaptation for such a complex life the lattice organs in having both a cuticular chamber and an cycle and may represent a synapomorphy for tantulocarids. In inner epidermals part, but in the tantulus, both chamber and these terms, we cannot compare tantulus and cypridoid instars. pore are devoid of any cellular processes, whereas the chamber But a cypridoid larva resembles sexual stages in generalized of lattice organs contains the paraciliary segment of the sensory Ascothoracida (Kolbasov et al. 2008;Høegetal.2009)ingen- cell. Moreover, lattice organs almost always occur in five pairs eral morphology, including the lattice organs. Therefore, in fur- in the dorsal surface, whereas there is no such fixed number ther studies, we are going to compare external morphology and among species of Tantulocarida. Lattice organs are considered anatomy of tantulocaridan sexual stages with cypridoid larvae one of the key morphological features assuring the monophyly to establish some possible homologies. of the Thecostraca and seem not to be purely larval structures as they are also present in adults of some Ascothoracida. It would Cement adhesion therefore be of high interest to search for possible homologies also in the sexual stages of the Tantulocarida. The Tantulocarida and the Cirripedia are unique in so far as adhesion is assured exclusively by means of chemical sub- Feeding adaptations stance produced in cephalic glands. In both the tantulus and the cypris of the Cirripedia, this proceeds by a secretion of The rootlet system of the Tantulocarida is obviously a unique cement from paired multicellular cement glands that exit first structure with few parallels in other parasitic crustaceans. through a cuticular lined secretion chamber and then through Outside the Thecostraca, the only example may be parasitic long cuticular ducts or canals (Walker 1971;Høeg1985). But copepods of the order Siphonostomatoida such as Rhizorhina there are also crucial differences. The cypris cement glands sp. (Kakui 2016). Within Thecostraca, so-called rootlet sys- exit onto each of the paired antennules whereas the four ducts tems developed three times, but clearly convergently and be- in the tantulus terminate onto the tip of the eversible and ing different in structure, ontogeny, and function. The best unpaired proboscis. Moreover, the secretion in the cypris is known example is the rootlets of the Cirripedia controlled by a muscular sac probably acting as a release valve Rhizocephala, which represents the whole body of the parasite whereas any such structure is absent in the tantulus. But at its endoparasitic stage (interna), and later provides nutrient 476 Petrunina A.S. et al. to the external reproductive sac. Two genera of parasitic References thoracican cirripedes, Rhizolepas infesting polychaetes and Anelasma infesting sharks, also have rootlets, but they are in Bielecki, J., Chan, B. K. K., Høeg, J. T., & Sari, A. (2009). Antennular both cases much less extensive and develop from the base of sensory organs in cyprids of balanomorphan cirripedes: standardiz- ing terminology using Megabalanus rosa. Biofouling, 25,203–214. the peduncle (Day 1939;Reesetal.2014). Boxshall, G. A. (1991). A review of the biology and phylogenetic relationships of the Tantulocarida, a subclass of Crustacea recognized in The stylet Cuticular stylets are developed in several groups of 1983. Verhandlungen der Deutschen Zoologischen Gesellschaft, 84, – parasitic crustaceans. Fish lice (Branchiura) have a so-called 271 279. Boxshall, G. A., & Lincoln, R. J. (1983). Tantulocarida, a new class of pre-oral stylet, an unpaired cuticular organ with two subtermi- Crustacea ectoparasitic on other crustaceans. Journal of Crustacean nal openings as part of their mouth apparatus. This stylet is Biology, 3(1), 1–16. supposedly used for injection of toxins into host tissues, but Boxshall, G. A., & Vader, W. (1993). A new genus of Tantulocarida not for obtaining nutrients (Walker et al. 2011). A much more (Crustacea) parasitic on an amphipod host from the North Sea. – relevant comparison is with the Cirripedia Rhizocephala. In Journal of Natural History, 27,977 988. Bresciani, J., & Høeg, J. T. (2001). Comparative ultrastructure of the root these, the kentrogon larva has a stylet that is used to clear the system in rhizocephalan barnacles (Crustacea: Cirripedia: way into host tissues for the invasive stage called the Rhizocephala). Journal of Morphology, 249,9–42. vermigon. Having few morphological characters, the stylet Day, J. H. (1939). A new cirripede parasite—Rhizolepas in rhizocephalans looks rather similar to that seen in the annelidicola,nov.gen.etsp.Proceedings Linnean Society – Tantulocarida, but they differ in both function, development, London, 151(2), 64 79. Glenner, H. (2001). Cypris metamorphosis, injection and earliest internal and the mechanism of penetrating into the host. In development of the rhizocephalan Loxothylacus panopaei (Gissler). rhizocephalans, the stylet is hollow from the base to the tip Crustacea: Cirripedia: Rhizocephala: Sacculinidae. Journal of and used not only to penetrate the host cuticle but also as a Morphology, 249,43–75. passageway for injection of the so-called vermigon larva as an Glenner, H., Høeg, J. T., Grygier, M. J., & Fujita, Y. (2008). Induced metamorphosis in crustacean y-larvae: towards a solution to a 100- endoparasite into the hemocoel of the host. Moreover, muscles year-old riddle. BMC Biology, 6(21), 1–6. are not involved in the penetration of the stylet of the Grygier, M. J. (1987). New records, external and internal anatomy, and kentrogon (Høeg 1985;Glenner2001;Høegetal.2012). In systematic position of Hansen’s y-larvae (Crustacea: Maxillopoda: tantulocarids, the stylet is operated by muscles and serves for Facetotecta). Sarsia, 72,261–278. punching a hole in the host cuticle which enables the rootlet Halcrow, K. (1993). Pore canal systems and associated microcuticular structures in amphipod crustaceans. In M. N. Horst & J. A. system to grow into the host. Moreover, the stylet is present Freeman (Eds.), The crustacean integument: morphology and already in the free-swimming tantulus larva, whereas in biochemistry (pp. 39–75). USA: CRC Press. rhizocephalans, it is only formed after settlement on the host. Hallberg, E., & Skog, M. (2011). Chemosensory sensilla in crustaceans. In T. Breithaupt & M. Thiel (Eds.), Chemical communication in crustaceans (pp. 103–121). New York: Springer. Higgins, R. P. (1964). A method for meiobenthic invertebrate collection. American Zoologist, 4,291. Conclusion Høeg, J. T. (1982). The anatomy and development of the rhizocephalan barnacle Clistosaccus paguri Lilljeborg and relation to its host The problem of position of Tantulocarida within Crustacea is Pagurus bernhardus (L.). Journal of Experimental Marine Biology and Ecology, 58(1), 87–125. still wide open. Since the tantulocarids are possibly nested Høeg, J. T. (1985). Cypris settlement, kentrogon formation and host in- within the monophyletic Thecostraca, there is a need for more vasion in the parasitic barnacle Lernaeodiscus porcellanae (Muller) molecular analysis and detailed comparison of different life (Crustacea: Cirripedia: Rhizocephala). Acta Zoologica (Stockholm), cycle stages. Particular emphasis should be put on the possible 66(1), 1–45. homology of the tantulus and the cypris larva and especially Høeg, J. T., & Kolbasov, G. A. (2002). Lattice organs in y-cyprids of the Facetotecta and their significance in the phylogeny of the Crustacea on comparison of their cement glands which despite obvious Thecostraca. Acta Zoologica, 83(1), 67–79. differences might still be homologous. Ultimately, there may Høeg, J. T., & Lützen, J. (1995). Life cycle and reproduction in the be a need to redefine the morphological characters diagnosing Cirripedia Rhizocephala. Oceanography and Marine Biology the Thecostraca in case the Tantulocarida prove to be a mem- Annual Review, 33,427–485. ber of this taxon. Høeg, J., Hosfeld, B., & Jensen, P. (1998). TEM studies on the lattice organs of cirripede cypris larvae (Crustacea, Thecostraca, Cirripedia). Zoomorphology, 118(4), 195–205. Acknowledgements The authors would like to thank Dr. Terue Kihara Høeg, J. T., Lagersson, N. C., & Glenner, H. (2004). The complete and Prof. Pedro Martinez Arbizu for providing facilities and assistance cypris larva and its significance in thecostracan phylogeny. In G. with the CLSM studies. For A.S.P. and G.A.K., this work was financially Scholtz (Ed.), Evolutionary and developmental biology of supported by the Russian Foundation for Basic Research (grants 17-54- Crustacea. Crustacean Issues 15 (pp. 197–215). Abingdon: 52006 MNT_a, 15-29-02447 ofi_m). Material sampling, SEM and TEM A.A. Balkema/Lisse. studies, and drafting of the paper were supported by the Russian Scientific Høeg, J. T., Pérez-Losada, M., Glenner, H., Kolbasov, G. A., & Crandall, Foundation, grant 14-50-00029. K. A. (2009). Evolution of morphology, ontogeny and life cycles Anatomy of the Tantulocarida: first results obtained using TEM and CLSM. Part I: tantulus larva 477

within the Crustacea Thecostraca. Arthropod Systematics and adult males (Crustacea, Thecostraca, Ascothoracida). Zoologischer Phylogeny, 67,199–217. Anzeiger, 247(3), 159–183. Høeg, J. T., Maruzzo, D., Okano, K., Glenner, H., & Chan, B. K. K. Makarova, A. A., & Polilov, A. A. (2013). Peculiarities of the brain (2012). Metamorphosis in balanomorphan, pedunculated, and para- organization and fine structure in small insects related to miniaturi- sitic barnacles: a video-based analysis. Integrative and Comparative zation. 1. The smallest Coleoptera (Ptiliidae). Entomological Biology, 52(3), 337–347. Review, 93(6), 703–713. Huys, R. (1989). Dicrotrichura tricincta gen. et spec. nov.: a new Martinez Arbizu, P., & Petrunina, A. S. (2017). Two new species of tantulocaridan (Crustacea: Maxillopoda) from Mediterranean deep Tantulocarida from the Atlantic deep sea with first CLSM pictures waters off Corsica. Bijdragen tot de Dierkunde, 59(4), 243–249. of tantulus larva. Mar Biodiversity. https://doi.org/10.1007/s12526- Huys, R. (1990). Campiloxiphos dineti gen. et spec. nov. from off 016-0627-6. Namibia and a redefinition of the Deoterthridae Boxshall and Huys R., Olesen, J., Petrunina, A.S., Martin, J.W., (2014). Tantulocarida. Lincoln (Crustacea: Tantulocarida). Journal of Natural History, In: Martin, J., Olesen, J., and Hoeg, J.T. (Eds) Atlas of crustacean – 24,415–432. larvae.(pp.122 127). Johns Hopkins University Press. Huys, R. (1991). Tantulocarida (Crustacea: Maxillopoda): a new taxon Pérez-Losada, M., Høeg, J. T., & Crandall, K. A. (2004). Unraveling the from the temporary meiobenthos. Pubblicazioni della Stazione evolutionary radiation of the thoracican barnacles using molecular Zoologica di Napoli I: Marine Ecology, 12(1), 1–34. and morphological evidence: a comparison of several divergence time estimation approaches. Systematic Biology, 53(2), 244–264. Huys, R., Andersen, P. F., & Kristensen, R. M. (1992). Tantulacus hoegi Pérez-Losada M., Høeg J.T., Crandall K.A. (2009). Remarkable conver- gen. et sp. nov. (Tantulocarida: Deoterthridae) from the meiobenthos gent evolution in specialized parasitic Thecostraca (Crustacea). of the Faroe Bank, North Atlantic. Sarsia, 76,287–297. BioMed Central Biology 7(15). https://doi.org/10.1186/1741-7007- Huys,R.,Boxshall,G.A.,&Lincoln,R.J.(1993).Thetantulocaridanlife 7-15. cycle: the circle closed? Journal of Crustacean Biology, 13(3), 432–442. Petrunina, A. S., Neretina, T. V.,Mugue, N. S., & Kolbasov, G. A. (2014). Jensen, P. G., Moyse, J., Høeg, J. T., & Al-Yahya, H. (1994). Tantulocarida vs Thecostraca: inside or outside. First attempts to Comparative SEM studies of lattice organs: putative sensory struc- resolve phylogenetic position of this parasitic taxon using gene se- tures on the carapace of larvae from Ascothoracida and Cirripedia quences. Journal of Zoological Systematics and Evolutionary (Crustacea Maxillopoda Thecostraca). Acta Zoologica (Stockholm), Research, 52,100–108. – 75,124 142. Rees, D. J., Noever, C., Høeg, J. T., Ommundsen, A., & Glenner, H. Kakui, K. (2016). Descriptions of two new species of Rhizorhina Hansen, (2014). On the origin of a novel parasitic-feeding mode within 1892 (Copepoda: Siphonostomatoida: Nicothoidae) parasitic on suspension-feeding barnacles. Current Biology, 24,1429–1434. tanaidacean crustaceans, with a note on their phylogenetic position. Savchenko, A.S., Kolbasov, G.A. (2009). Serratotantulus chertoprudae – Systematic parasitology, 93,57 68. gen. et sp. n. (Crustacea, Tantulocarida, Basipodellidae)—anew Khodami, S., McArthur, J. V., Blanco-Bercial, L., & Martinez Arbizu, P. tantulocaridan from the abyssal depths of the Indian Ocean. (2017). Molecular phylogeny and revision of copepod orders Integrative and Comparative Biology, Symposium BThe Biology (Crustacea: Copepoda). Scientific Reports, 7, 9164. of the Parasitic Crustacea^,January3–7, 2009. Boston, Kolbasov, G. A., & Høeg, J. T. (2007). Cypris larvae of acrothoracican Massachusetts. P. 1–8. barnacles (Thecostraca: Cirripedia: Acrothoracica). Zoologischer Walker, G. (1971). A study of the cement apparatus of the cypris larva of Anzeiger, 6(2), 127–151. the barnacle Balanus balanoides. Marine Biology, 9,205–212. Kolbasov, G. A., Grygier, M. J., Høeg, J. T., & Klepal, W. (2008). Walker, P. D., Russon, I. J., Haond, C., Van der Velde, G., & Bonga, S. E. External morphology of the two cypridiform ascothoracid-larva in- W. (2011). Feeding in adult Argulus japonicus Thiele, 1900 stars of Dendrogaster: the evolutionary significance of the two-step (Maxillopoda, Branchiura), an ectoparasite on fish. Crustaceana, metamorphosis and comparison of lattice organs between larvae and 84(3), 307–318.